red-datasets 0.0.7 → 0.0.8

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 222271b814e3a5ce23b5e0dd1d2578bffb84afdab10110b0869985c6056bfd3b
-  data.tar.gz: ac30931b3317ab04afd394b28a45a9206c784d78b3bcaf98fc3a2a48227c7930
+  metadata.gz: c7a9199546e7a001c97e45c6fa28db15c0d96b748e527d9705dfee4e4b1db6fd
+  data.tar.gz: c659f6ae1e658ad91210e4427be063463124d89ef90388d34ebfb73ceb49068a
 SHA512:
-  metadata.gz: 8a94a3d66baaed4948904e97dc53100d73ae96c528c09b02252caabd05b8545587abf6fbcba3a578725812327a9a2c8827bbb7e283ccd3d7e66753bf30035e2e
-  data.tar.gz: 2ab44b5aa3ee5da0ac8e8307546c71942938de4497bfec05fc929715a4e5ef6df1cb091bce0d5f12978582d2c9fa7eaffff9edd54be0d845627dccfce42a63dd
+  metadata.gz: d8a23c4a165a596df22ce5bbe1f8f0cd5c0f002deecafbb26cd5e5f75abb3c0224c1013898162a67787159258d1b801395fc4d949c17939d95940664cffd5600
+  data.tar.gz: f2fd4eb733e6205f138c4005627e815e3787040a8a4b6cce7eca9fd5d4adaa12263e17e8f5bd9394a851e5210f28736ee3c682c81e110da304ae17fb3f0bedba

data/README.md

@@ -1,8 +1,4 @@
-# README
-
-## Name
-
-Red Datasets
+# Red Datasets
 
 ## Description
 
@@ -16,6 +12,20 @@ You can use datasets easily because you can access each dataset with multiple wa
 % gem install red-datasets
 ```
 
+## Available datasets
+
+TODO: Document them in source code to list in document: https://www.rubydoc.info/gems/red-datasets
+
+* Adult Dataset
+* CIFAR-10 Dataset
+* CIFAR-100 Dataset
+* Fashion-MNIST
+* Iris Dataset
+* MNIST database
+* The Penn Treebank Project
+* Wikipedia
+* Wine Dataset
+
 ## Usage
 
 Here is an example to access [Iris Data Set](https://archive.ics.uci.edu/ml/datasets/iris) by `#each`  or `Table#to_h` or `Table#fetch_values`.

data/doc/text/news.md

@@ -1,5 +1,22 @@
 # News
 
+## 0.0.8 - 2019-03-24
+
+### Improvements
+
+  * Improved README.
+    [GitHub#40][Patch by kojix2]
+
+  * `Datasets::PostalCodeJapan`: Added.
+
+  * `Datasets::LIBSVMDatasetList`: Added.
+
+  * `Datasets::LIBSVM`: Added.
+
+### Thanks
+
+  * kojix2
+
 ## 0.0.7 - 2018-11-21
 
 ### Improvements

data/lib/datasets/libsvm-dataset-list.rb

@@ -0,0 +1,137 @@
+require "English"
+require "rexml/document"
+
+require_relative "dataset"
+
+module Datasets
+  class LIBSVMDatasetList < Dataset
+    File = Struct.new(:name,
+                      :url,
+                      :note)
+    class Record < Struct.new(:name,
+                              :source,
+                              :preprocessing,
+                              :n_classes,
+                              :n_data,
+                              :n_features,
+                              :files)
+      def to_h
+        hash = super
+        hash[:files] = hash[:files].collect(&:to_h)
+        hash
+      end
+    end
+
+    def initialize
+      super()
+      @metadata.id = "libsvm-dataset-list"
+      @metadata.name = "LIBSVM dataset list"
+      @metadata.url = "https://www.csie.ntu.edu.tw/~cjlin/libsvmtools/datasets/"
+      @metadata.description = lambda do
+        extract_description
+      end
+    end
+
+    def each
+      return to_enum(__method__) unless block_given?
+
+      open_data do |input|
+        # TODO: Improve performance
+        document = REXML::Document.new(input)
+        is_header = true
+        document.each_element("//tr") do |tr|
+          if is_header
+            is_header = false
+            next
+          end
+          name = tr.elements.first
+          a = name.elements.first
+          href = a.attributes["href"]
+          record = Record.new
+          record.name = a.text
+          record.files = []
+          parse_detail(href, record)
+          yield(record)
+        end
+      end
+    end
+
+    private
+    def open_data
+      data_path = cache_dir_path + "index.html"
+      unless data_path.exist?
+        download(data_path, @metadata.url)
+      end
+      ::File.open(data_path) do |input|
+        yield(input)
+      end
+    end
+
+    def extract_description
+      open_data do |input|
+        document = REXML::Document.new(input)
+        description = []
+        in_content = false
+        document.each_element("//body/*") do |element|
+          unless in_content
+            in_content = (element.name == "h1")
+            next
+          end
+          break if element.name == "hr"
+          content = extract_text(element)
+          description << content unless content.empty?
+        end
+        description.join("\n\n")
+      end
+    end
+
+    def extract_text(element)
+      texts = REXML::XPath.match(element, ".//text()")
+      texts.join("").gsub(/[ \t\n]+/, " ").strip
+    end
+
+    def open_detail(detail)
+      data_path = cache_dir_path + detail
+      unless data_path.exist?
+        download(data_path, @metadata.url + detail)
+      end
+      ::File.open(data_path) do |input|
+        yield(input)
+      end
+    end
+
+    def parse_detail(href, record)
+      path, id = href.split("#")
+      open_detail(path) do |detail|
+        detail_document = REXML::Document.new(detail)
+        anchor = REXML::XPath.match(detail_document, "//*[@name='#{id}']")[0]
+        ul = anchor.next_sibling
+        ul.each_element do |li|
+          text = extract_text(li)
+          case text
+          when /\ASource: /
+            record.source = $POSTMATCH
+          when /\APreprocessing: /
+            record.preprocessing = $POSTMATCH
+          when /\A\# of classes: (\d+)/
+            record.n_classes = Integer($1, 10)
+          when /\A\# of data: ([\d,]+)/
+            record.n_data = Integer($1.gsub(/,/, ""), 10)
+          when /\A\# of features: ([\d,]+)/
+            record.n_features = Integer($1.gsub(/,/, ""), 10)
+          when /\AFiles:/
+            li.elements.first.each_element do |file_li|
+              file_a = file_li.elements.first
+              file = File.new
+              file.name = file_a.text
+              file.url = @metadata.url + file_a.attributes["href"]
+              file_note = file_li.text
+              file.note = file_note.strip.gsub(/[()]/, "") if file_note
+              record.files << file
+            end
+          end
+        end
+      end
+    end
+  end
+end

data/lib/datasets/libsvm.rb

@@ -0,0 +1,143 @@
+require "csv"
+
+require_relative "dataset"
+
+module Datasets
+  class LIBSVM < Dataset
+    class Record
+      attr_reader :label
+      attr_reader :features
+      def initialize(label, features)
+        @label = label
+        @features = features
+      end
+
+      def [](index)
+        @features[index]
+      end
+
+      def to_h
+        hash = {
+          label: @label,
+        }
+        @features.each_with_index do |feature, i|
+          hash[i] = feature
+        end
+        hash
+      end
+
+      def values
+        [@label] + @features
+      end
+    end
+
+    def initialize(name,
+                   note: nil,
+                   default_feature_value: 0)
+      super()
+      @libsvm_dataset_metadata = fetch_dataset_info(name)
+      @file = choose_file(note)
+      @default_feature_value = default_feature_value
+      @metadata.id = "libsvm-#{normalize_name(name)}"
+      @metadata.name = "LIBSVM dataset: #{name}"
+      @metadata.url = "https://www.csie.ntu.edu.tw/~cjlin/libsvmtools/datasets/"
+    end
+
+    def each
+      return to_enum(__method__) unless block_given?
+
+      open_data do |input|
+        n_features = @libsvm_dataset_metadata.n_features
+        csv = CSV.new(input, col_sep: " ")
+        csv.each do |row|
+          label = parse_label(row.shift)
+          features = [@default_feature_value] * n_features
+          row.each do |column|
+            next if column.nil?
+            index, value = column.split(":", 2)
+            features[Integer(index, 10) - 1] = parse_value(value)
+          end
+          yield(Record.new(label, features))
+        end
+      end
+    end
+
+    private
+    def fetch_dataset_info(name)
+      list = LIBSVMDatasetList.new
+      available_datasets = []
+      list.each do |record|
+        available_datasets << record.name
+        if record.name == name
+          return record
+        end
+      end
+      message = "unavailable LIBSVM dataset: #{name.inspect}: "
+      message << "available datasets: ["
+      message << available_datasets.collect(&:inspect).join(", ")
+      message << "]"
+      raise ArgumentError, message
+    end
+
+    def choose_file(note)
+      files = @libsvm_dataset_metadata.files
+      return files.first if note.nil?
+
+      available_notes = []
+      @libsvm_dataset_metadata.files.find do |file|
+        return file if file.note == note
+        available_notes << file.note if file.note
+      end
+
+      name = @libsvm_dataset_metadata.name
+      message = "unavailable note: #{name}: #{note.inspect}: "
+      message << "available notes: ["
+      message << available_notes.collect(&:inspect).join(", ")
+      message << "]"
+      raise ArgumentError, message
+    end
+
+    def open_data(&block)
+      data_path = cache_dir_path + @file.name
+      unless data_path.exist?
+        download(data_path, @file.url)
+      end
+      if data_path.extname == ".bz2"
+        input, output = IO.pipe
+        pid = spawn("bzcat", data_path.to_s, {:out => output})
+        begin
+          output.close
+          yield(input)
+        ensure
+          input.close
+          Process.waitpid(pid)
+        end
+      else
+        File.open(data_path, &block)
+      end
+    end
+
+    def normalize_name(name)
+      name.gsub(/[()]/, "").gsub(/[ _;]+/, "-").downcase
+    end
+
+    def parse_label(label)
+      labels = label.split(",").collect do |value|
+        parse_value(value)
+      end
+      if labels.size == 1
+        labels[0]
+      else
+        labels
+      end
+    end
+
+    def parse_value(value)
+      if value.include?(".")
+        Float(value)
+      else
+        Integer(value, 10)
+      end
+    end
+  end
+end

data/lib/datasets/postal-code-japan.rb

@@ -0,0 +1,154 @@
+require "csv"
+require "zip"
+
+require_relative "dataset"
+
+module Datasets
+  class PostalCodeJapan < Dataset
+    class Record < Struct.new(:organization_code,
+                              :old_postal_code,
+                              :postal_code,
+                              :prefecture_reading,
+                              :city_reading,
+                              :address_reading,
+                              :prefecture,
+                              :city,
+                              :address,
+                              :have_multiple_postal_codes,
+                              :have_address_number_per_koaza,
+                              :have_chome,
+                              :postal_code_is_shared,
+                              :changed,
+                              :change_reason)
+      alias_method :have_multiple_postal_codes?,
+                   :have_multiple_postal_codes
+      alias_method :have_address_number_per_koaza?,
+                   :have_address_number_per_koaza
+      alias_method :have_chome?,
+                   :have_chome
+      alias_method :postal_code_is_shared?,
+                   :postal_code_is_shared
+      alias_method :changed?,
+                   :changed
+    end
+
+    VALID_READINGS = [
+      :lowercase,
+      :uppercase,
+      :romaji,
+    ]
+    def initialize(reading: :lowercase)
+      super()
+      @reading = reading
+      unless VALID_READINGS.include?(@reading)
+        message = ":reading must be one of ["
+        message << VALID_READINGS.collect(&:inspect).join(", ")
+        message << "]: #{@reading.inspect}"
+        raise ArgumentError, message
+      end
+      @metadata.id = "postal-code-japan-#{@reading}"
+      @metadata.name = "Postal code in Japan (#{@reading})"
+      @metadata.url = "https://www.post.japanpost.jp/zipcode/download.html"
+      @metadata.licenses = [
+        "CC0-1.0",
+      ]
+      @metadata.description = "Postal code in Japan (reading: #{@reading})"
+    end
+
+    def each(&block)
+      return to_enum(__method__) unless block_given?
+
+      open_data do |input|
+        utf8_data = input.read.encode(Encoding::UTF_8, Encoding::CP932)
+        options = {
+          quote_char: nil,
+          strip: %Q["],
+        }
+        if @reading == :romaji
+          CSV.parse(utf8_data, **options) do |row|
+            yield(Record.new(nil,
+                             nil,
+                             row[0],
+                             row[4],
+                             row[5],
+                             row[6],
+                             row[1],
+                             row[2],
+                             row[3],
+                             false,
+                             false,
+                             false,
+                             false,
+                             false,
+                             nil))
+          end
+        else
+          CSV.parse(utf8_data, **options) do |row|
+            yield(Record.new(row[0],
+                             row[1].rstrip,
+                             row[2],
+                             row[3],
+                             row[4],
+                             row[5],
+                             row[6],
+                             row[7],
+                             row[8],
+                             (row[9] == "1"),
+                             (row[10] == "1"),
+                             (row[11] == "1"),
+                             (row[12] == "1"),
+                             (row[13] != "0"),
+                             convert_change_reason(row[14])))
+          end
+        end
+      end
+    end
+
+    private
+    def open_data
+      data_url = "https://www.post.japanpost.jp/zipcode/dl"
+      case @reading
+      when :lowercase
+        data_url << "/kogaki/zip/ken_all.zip"
+      when :uppercase
+        data_url << "/oogaki/zip/ken_all.zip"
+      when :romaji
+        data_url << "/roman/ken_all_rome.zip"
+      end
+      data_path = cache_dir_path + "#{@reading}-ken-all.zip"
+      unless data_path.exist?
+        download(data_path, data_url)
+      end
+
+      Zip::File.open(data_path.to_s) do |zip_file|
+        zip_file.each do |entry|
+          next unless entry.file?
+          entry.get_input_stream do |input|
+            yield(input)
+          end
+        end
+      end
+    end
+
+    def convert_change_reason(reason)
+      case reason
+      when "0"
+        nil
+      when "1"
+        :new
+      when "2"
+        :japanese_addressing_system
+      when "3"
+        :land_readjustment
+      when "4"
+        :postal_district_adjustment
+      when "5"
+        :correction
+      when "6"
+        :deletion
+      else
+        :unknown
+      end
+    end
+  end
+end

data/lib/datasets/version.rb

@@ -1,3 +1,3 @@
 module Datasets
-  VERSION = "0.0.7"
+  VERSION = "0.0.8"
 end

data/lib/datasets.rb

@@ -1,10 +1,13 @@
-require "datasets/version"
+require_relative "datasets/version"
 
-require "datasets/adult"
-require "datasets/cifar"
-require "datasets/fashion-mnist"
-require "datasets/iris"
-require "datasets/mnist"
-require "datasets/penn-treebank"
-require "datasets/wikipedia"
-require "datasets/wine"
+require_relative "datasets/adult"
+require_relative "datasets/cifar"
+require_relative "datasets/fashion-mnist"
+require_relative "datasets/iris"
+require_relative "datasets/libsvm"
+require_relative "datasets/libsvm-dataset-list"
+require_relative "datasets/mnist"
+require_relative "datasets/penn-treebank"
+require_relative "datasets/postal-code-japan"
+require_relative "datasets/wikipedia"
+require_relative "datasets/wine"

data/red-datasets.gemspec

@@ -34,6 +34,9 @@ Gem::Specification.new do |spec|
   spec.files += Dir.glob("doc/text/*")
   spec.test_files += Dir.glob("test/**/*")
 
+  spec.add_runtime_dependency("csv", ">= 3.0.5")
+  spec.add_runtime_dependency("rubyzip")
+
   spec.add_development_dependency("bundler")
   spec.add_development_dependency("rake")
   spec.add_development_dependency("test-unit")

data/test/test-libsvm-dataset-list.rb

@@ -0,0 +1,47 @@
+class LIBSVMDatasetListTest < Test::Unit::TestCase
+  def setup
+    @dataset = Datasets::LIBSVMDatasetList.new
+  end
+
+  test("#each") do
+    assert_equal({
+                   name: "a1a",
+                   source: "UCI / Adult",
+                   preprocessing:
+                     "The original Adult data set has 14 features, " +
+                     "among which six are continuous and eight are " +
+                     "categorical. In this data set, continuous features " +
+                     "are discretized into quantiles, and each quantile is " +
+                     "represented by a binary feature. Also, a categorical " +
+                     "feature with m categories is converted to m binary " +
+                     "features. Details on how each feature is converted " +
+                     "can be found in the beginning of each file from this " +
+                     "page. [JP98a]",
+                   n_classes: 2,
+                   n_data: 1605,
+                   n_features: 123,
+                   files: [
+                     {
+                       name: "a1a",
+                       url: "https://www.csie.ntu.edu.tw/~cjlin/libsvmtools/datasets/binary/a1a",
+                       note: nil,
+                     },
+                     {
+                       name: "a1a.t",
+                       url: "https://www.csie.ntu.edu.tw/~cjlin/libsvmtools/datasets/binary/a1a.t",
+                       note: "testing",
+                     }
+                   ],
+                 },
+                 @dataset.first.to_h)
+  end
+
+  sub_test_case("#metadata") do
+    test("#description") do
+      description = @dataset.metadata.description
+      assert do
+        description.start_with?("This page contains many classification, ")
+      end
+    end
+  end
+end

data/test/test-libsvm.rb

@@ -0,0 +1,205 @@
+class LIBSVMDatasetTest < Test::Unit::TestCase
+  test(":note") do
+    dataset = Datasets::LIBSVM.new("a1a", note: "testing")
+    hash = {label: -1}
+    n_features = 123
+    n_features.times do |i|
+      hash[i] = 0
+    end
+    [5, 7, 14, 19, 39, 40, 51, 63, 67, 73, 74, 76, 78, 83].each do |i|
+      hash[i - 1] = 1
+    end
+    assert_equal(hash,
+                 dataset.first.to_h)
+  end
+
+  test(":default_feature_value") do
+    dataset = Datasets::LIBSVM.new("a1a", default_feature_value: nil)
+    hash = {label: -1}
+    n_features = 123
+    n_features.times do |i|
+      hash[i] = nil
+    end
+    [3, 11, 14, 19, 39, 42, 55, 64, 67, 73, 75, 76, 80, 83].each do |i|
+      hash[i - 1] = 1
+    end
+    assert_equal(hash,
+                 dataset.first.to_h)
+  end
+
+  test("classification") do
+    dataset = Datasets::LIBSVM.new("a1a")
+    hash = {label: -1}
+    n_features = 123
+    n_features.times do |i|
+      hash[i] = 0
+    end
+    [3, 11, 14, 19, 39, 42, 55, 64, 67, 73, 75, 76, 80, 83].each do |i|
+      hash[i - 1] = 1
+    end
+    assert_equal(hash,
+                 dataset.first.to_h)
+  end
+
+  test("regression") do
+    dataset = Datasets::LIBSVM.new("abalone")
+    hash = {label: 15}
+    n_features = 8
+    n_features.times do |i|
+      hash[i] = 0
+    end
+    [
+      [1, 1],
+      [2, 0.455],
+      [3, 0.365],
+      [4, 0.095],
+      [5, 0.514],
+      [6, 0.2245],
+      [7, 0.101],
+      [8, 0.15],
+    ].each do |i, value|
+      hash[i - 1] = value
+    end
+    assert_equal(hash,
+                 dataset.first.to_h)
+  end
+
+  test("multi-label") do
+    dataset = Datasets::LIBSVM.new("mediamill (exp1)")
+    hash = {label: [65, 67, 11, 31]}
+    n_features = 120
+    n_features.times do |i|
+      hash[i] = 0
+    end
+    [
+      [1, 0.380877],
+      [2, 0.494079],
+      [3, 0.540009],
+      [4, 0.422926],
+      [5, 0.158318],
+      [6, 0.326975],
+      [7, 0.390861],
+      [8, 0.527121],
+      [9, 0.254052],
+      [10, 0.223731],
+      [11, 0.040285],
+      [12, 0.141133],
+      [13, 0.112249],
+      [14, 0.263171],
+      [15, 0.147020],
+      [16, 0.472414],
+      [17, 0.592614],
+      [18, 0.653138],
+      [19, 0.499867],
+      [20, 0.196520],
+      [21, 0.403892],
+      [22, 0.482395],
+      [23, 0.619219],
+      [24, 0.320346],
+      [25, 0.281251],
+      [26, 0.054750],
+      [27, 0.180459],
+      [28, 0.139964],
+      [29, 0.319925],
+      [30, 0.181216],
+      [31, 0.364294],
+      [32, 0.407211],
+      [33, 0.368926],
+      [34, 0.427661],
+      [35, 0.211391],
+      [36, 0.364345],
+      [37, 0.370710],
+      [38, 0.409107],
+      [39, 0.289299],
+      [40, 0.243053],
+      [41, 0.063121],
+      [42, 0.193587],
+      [43, 0.158755],
+      [44, 0.316054],
+      [45, 0.197410],
+      [46, 0.656168],
+      [47, 0.678760],
+      [48, 0.650831],
+      [49, 0.674636],
+      [50, 0.492428],
+      [51, 0.623887],
+      [52, 0.610622],
+      [53, 0.678219],
+      [54, 0.574774],
+      [55, 0.523073],
+      [56, 0.206804],
+      [57, 0.496294],
+      [58, 0.429221],
+      [59, 0.586611],
+      [60, 0.471550],
+      [61, 0.284480],
+      [62, 0.432466],
+      [63, 0.498075],
+      [64, 0.408141],
+      [65, 0.102713],
+      [66, 0.303028],
+      [67, 0.309501],
+      [68, 0.444855],
+      [69, 0.191727],
+      [70, 0.174895],
+      [71, 0.034143],
+      [72, 0.153099],
+      [73, 0.068318],
+      [74, 0.217020],
+      [75, 0.099688],
+      [76, 0.409862],
+      [77, 0.561918],
+      [78, 0.612031],
+      [79, 0.514471],
+      [80, 0.146015],
+      [81, 0.398807],
+      [82, 0.383295],
+      [83, 0.548485],
+      [84, 0.282937],
+      [85, 0.252712],
+      [86, 0.051008],
+      [87, 0.223110],
+      [88, 0.098112],
+      [89, 0.299672],
+      [90, 0.144873],
+      [91, 0.308488],
+      [92, 0.358478],
+      [93, 0.352077],
+      [94, 0.394686],
+      [95, 0.157513],
+      [96, 0.339370],
+      [97, 0.321558],
+      [98, 0.341373],
+      [99, 0.247969],
+      [100, 0.206070],
+      [101, 0.061001],
+      [102, 0.216793],
+      [103, 0.112389],
+      [104, 0.273648],
+      [105, 0.152745],
+      [106, 0.598081],
+      [107, 0.621687],
+      [108, 0.607213],
+      [109, 0.644025],
+      [110, 0.394948],
+      [111, 0.593651],
+      [112, 0.551529],
+      [113, 0.574392],
+      [114, 0.511032],
+      [115, 0.463997],
+      [116, 0.202034],
+      [117, 0.492341],
+      [118, 0.317983],
+      [119, 0.547807],
+      [120, 0.393778],
+    ].each do |i, value|
+      hash[i - 1] = value
+    end
+    assert_equal(hash,
+                 dataset.first.to_h)
+  end
+
+  test("string") do
+    # TODO
+  end
+end

data/test/test-postal-code-japan.rb

@@ -0,0 +1,69 @@
+class PostalCodeJapanTest < Test::Unit::TestCase
+  sub_test_case(":reading") do
+    test(":lowercase") do
+      dataset = Datasets::PostalCodeJapan.new(reading: :lowercase)
+      assert_equal({
+                     organization_code: "01101",
+                     old_postal_code: "060",
+                     postal_code: "0600000",
+                     prefecture_reading: "ﾎｯｶｲﾄﾞｳ",
+                     city_reading: "ｻｯﾎﾟﾛｼﾁｭｳｵｳｸ",
+                     address_reading: "ｲｶﾆｹｲｻｲｶﾞﾅｲﾊﾞｱｲ",
+                     prefecture: "北海道",
+                     city: "札幌市中央区",
+                     address: "以下に掲載がない場合",
+                     have_multiple_postal_codes: false,
+                     have_address_number_per_koaza: false,
+                     have_chome: false,
+                     postal_code_is_shared: false,
+                     changed: false,
+                     change_reason: nil,
+                   },
+                   dataset.first.to_h)
+    end
+
+    test(":uppercase") do
+      dataset = Datasets::PostalCodeJapan.new(reading: :uppercase)
+      assert_equal({
+                     organization_code: "01101",
+                     old_postal_code: "060",
+                     postal_code: "0600000",
+                     prefecture_reading: "ﾎﾂｶｲﾄﾞｳ",
+                     city_reading: "ｻﾂﾎﾟﾛｼﾁﾕｳｵｳｸ",
+                     address_reading: "ｲｶﾆｹｲｻｲｶﾞﾅｲﾊﾞｱｲ",
+                     prefecture: "北海道",
+                     city: "札幌市中央区",
+                     address: "以下に掲載がない場合",
+                     have_multiple_postal_codes: false,
+                     have_address_number_per_koaza: false,
+                     have_chome: false,
+                     postal_code_is_shared: false,
+                     changed: false,
+                     change_reason: nil,
+                   },
+                   dataset.first.to_h)
+    end
+
+    test(":romaji") do
+      dataset = Datasets::PostalCodeJapan.new(reading: :romaji)
+      assert_equal({
+                     organization_code: nil,
+                     old_postal_code: nil,
+                     postal_code: "0600000",
+                     prefecture_reading: "HOKKAIDO",
+                     city_reading: "SAPPORO SHI CHUO KU",
+                     address_reading: "IKANIKEISAIGANAIBAAI",
+                     prefecture: "北海道",
+                     city: "札幌市　中央区",
+                     address: "以下に掲載がない場合",
+                     have_multiple_postal_codes: false,
+                     have_address_number_per_koaza: false,
+                     have_chome: false,
+                     postal_code_is_shared: false,
+                     changed: false,
+                     change_reason: nil,
+                   },
+                   dataset.first.to_h)
+    end
+  end
+end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: red-datasets
 version: !ruby/object:Gem::Version
-  version: 0.0.7
+  version: 0.0.8
 platform: ruby
 authors:
 - tomisuker
@@ -9,8 +9,36 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2018-11-20 00:00:00.000000000 Z
+date: 2019-03-24 00:00:00.000000000 Z
 dependencies:
+- !ruby/object:Gem::Dependency
+  name: csv
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 3.0.5
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 3.0.5
+- !ruby/object:Gem::Dependency
+  name: rubyzip
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
 - !ruby/object:Gem::Dependency
   name: bundler
   requirement: !ruby/object:Gem::Requirement
@@ -106,9 +134,12 @@ files:
 - lib/datasets/downloader.rb
 - lib/datasets/fashion-mnist.rb
 - lib/datasets/iris.rb
+- lib/datasets/libsvm-dataset-list.rb
+- lib/datasets/libsvm.rb
 - lib/datasets/metadata.rb
 - lib/datasets/mnist.rb
 - lib/datasets/penn-treebank.rb
+- lib/datasets/postal-code-japan.rb
 - lib/datasets/table.rb
 - lib/datasets/version.rb
 - lib/datasets/wikipedia.rb
@@ -121,8 +152,11 @@ files:
 - test/test-dictionary.rb
 - test/test-fashion-mnist.rb
 - test/test-iris.rb
+- test/test-libsvm-dataset-list.rb
+- test/test-libsvm.rb
 - test/test-mnist.rb
 - test/test-penn-treebank.rb
+- test/test-postal-code-japan.rb
 - test/test-table.rb
 - test/test-wikipedia.rb
 - test/test-wine.rb
@@ -146,20 +180,23 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 3.0.0.beta2
+rubygems_version: 2.7.6
 signing_key: 
 specification_version: 4
 summary: Red Datasets provides classes that provide common datasets such as iris dataset.
 test_files:
-- test/test-wine.rb
-- test/run-test.rb
-- test/test-cifar.rb
-- test/test-fashion-mnist.rb
-- test/test-wikipedia.rb
 - test/test-iris.rb
-- test/helper.rb
+- test/test-wikipedia.rb
+- test/test-fashion-mnist.rb
+- test/test-wine.rb
+- test/test-postal-code-japan.rb
 - test/test-mnist.rb
-- test/test-table.rb
+- test/helper.rb
 - test/test-adult.rb
+- test/test-libsvm.rb
+- test/run-test.rb
+- test/test-table.rb
+- test/test-cifar.rb
+- test/test-libsvm-dataset-list.rb
 - test/test-penn-treebank.rb
 - test/test-dictionary.rb